The vertebral column and skeletal musculature derive from embryonic structures called somites, which form sequentially from head to tail. Normal musculoskeletal development requires the correct number of cells in each somite and that each somite acquires its correct axial address. Perturbations can lead to malformations of the spine ranging from complete disarray of the vertebral elements (e.g. spondylocostal and spondylothoracic dysostoses and dysplasias) to deviations of the spine (lordosis, kyphosis, and scoliosis) and misspecification of the regional identity of skeletal elements (e.g. Klippel-Feil syndrome). Research in the last two decades has uncovered molecular oscillators and gradients of growth factors hypothesized to control somite size and identity. However, little is known about how (or if) these molecular players relate to cell behaviors like cell adhesion, proliferation and migration that result in somite patterning. In part this uncertainty is due to the spatiotemporal complexity of somitogenesis, the number of mechanisms involved and the relative lack of cross-talk between model and experiment in the past. This project undertakes a multiscale approach to address all three issues. The NIH-led Interagency Modeling and Analysis Group (IMAG) has identified as key goals the development of open-source, multi-scale biological simulation environments and the deployment of demonstration projects that integrate models operating at different scales. This project will build comprehensive 3D multiscale predictive models of vertebrate somitogenesis able to generate and test specific hypotheses concerning the mechanisms of interspecies differences (as a model of individual to individual variability and robustness) and perturbations. It will refine a tissue simulation environment (CC3D) to improve its usability to the community, perform new biological experiments to collect data as inputs for 3D somitogenesis models and to test model predictions, and deploy models and experimental data using emerging standards for sharing of multicellular information (CBO, CBMSL). Specifically, it will: 1) develop new 3D models to integrate behaviors at molecular, cellular and tissue scales to reproduce the normal dynamics of segmentation and test them by quantitative measurements using advanced time-lapse fluorescence microscopy and microfluidics-based gradient-cell technology; 2) use a novel experimental paradigm that allows segmentation to be studied independently of the molecular segmentation clock, for challenging and validating the segmentation models; 3) extend the models and experiments to understand how somites acquire positional identities and 4) open-source release data and models in sharable formats. In addition to generating a predictive model for vertebral column development and its anomalies, this project should enable future studies of the development of other organs and establish the role of multi-scale modeling in biomedical science. Its emphasis on model and data share ability will promote efficient sharing of resources, tools and models among biomodelers and experimentalists, significantly reducing duplication of effort.